2017年度数学与统计学院校内专家学术报告两则
时间:2017-03-26 15:14:55 来源: 作者:数学学院 阅读: 次
报告摘要: The multi-innovation identification method extends the concept of the single innovation identification method. By introducing the innovation length and expanding the dimension of the innovation, the multi-innovation identification method fully utilizes the characteristic that the innovation can improve the accuracy of the parameter estimates and the state estimates. In this talk, we discuss the multi-innovation identification methods for three typical systems, including Box-Jenkins systems, the bilinear-in-parameter systems and the input nonlinear state space systems with colored noise. The simulation examples are given for each proposed algorithm to test the effectiveness. The convergence for some important identification algorithms is analyzed theoretically.